Understanding the basis of disease and the development of therapeutic and preventative treatments has evolved over the last century from empirical observation and experimentation to genome wide mutation scanning. The revolution in genomics has provided researchers with the tools to look for a genomic basis for disease. The Human Genome effort has generated a raw sequence of the 3 billion base pairs of the human genome and revealed about 35,000 genes. Genetic variations amongst different individuals and in and in between populations are being studied in order to determine the association with the predisposition to disease or the correlation to drug efficacy and/or side effects. The promise of personalized medicine based on a panel of genetic markers has tantalized the healthcare community and provides an important goal for those focused on providing diagnostic and treatment options for healthcare providers and patients.
With the development of a variety of tools in molecular biology, such as nucleic amplification methods, cloning and expression systems and methods, disease analysis has been based on a genomics, or bottom up, approach. This approach presumes that a genetic change or set of changes will have a long reaching effect on protein function by affecting mRNA transcription or protein structure and function.
Technologies have been developed to analyze single nucleotide polymorphisms (SNPs) in an industrial scale (e.g., MassARRAY™ and the MassARRAY® system, Sequenom, Inc., San Diego, Calif.) and in pooled samples to study the frequency of SNPs in populations of various gender, ethnicity, age and health condition. The ultimate goal of these efforts is to understand the etiology of disease on the molecular level (e.g., based on genetic variances (pharmacogenomics)), to develop diagnostic assays and effective drugs with few or no side-effects.
Genomics has fallen short of the original expectation that this strategy could be used to stratify a population relative to a defined phenotype, including differences between normal and disease patient population or populations. Although single genetic markers have been found to be associated with or cause or predict a specific disease state, genomic information may not be sufficient to stratify individual populations by of the association of an SNP (or SNPs) with a given disease, drug side-effect or other target phenotype. Because of the large number of potential targets and regulatory signals that affect protein translation, it is not sufficient to establish the differential expression profiles of messenger RNA in comparing phenotypes or populations, such as healthy and disease states, or such as the analyses using expression DNA chips (e.g., GeneChip™ technology, Affymetrix, Inc., Santa Clara, Calif.; LifeArray™ technology, Incyte Genomics, Inc., Palo Alto, Calif.). The metabolic activities in a cell are not performed by mRNA but rather by the translated proteins and subsequently posttranslationally modified products, such as the alkylated, glycosylated and phosphorylated products.
The study of proteomics encompasses the study of individual proteins and how these proteins function within a biochemical pathway. Proteomics also includes the study of protein interactions, including how they form the architecture that constitutes living cells. In many human diseases such as cancer, Alzheimer's disease, diabetes as well as host responses to infectious diseases, the elucidation of the complex interactions between regulatory proteins, which can cause diseases, is a critical step to finding effective treatment. Often, SNPs and other nucleic acid mutations occur in genes whose products are such proteins as (1) growth related hormones, (2) membrane receptors for growth hormones, (3) components of the trans-membrane signal pathway and (4) DNA binding proteins that act on transcription and the inactivation of suppressor genes (e.g. p53) causing the onset of disease.
Complex protein mixtures are analyzed by two-dimensional (2D) gel electrophoresis and subsequent image processing to identify changes in the pattern (structural changes) or intensity of various protein spots. Two-dimensional gel electrophoresis is a laborious, error-prone method with low reproducibility and cannot be effectively automated. This gel technology is unable to effectively analyze membrane proteins. Further, the resolution of 2D gels is insufficient to analyze the profile of all proteins present in a mixture.
Available protein chips are limited by their ability to specifically capture hydrophobic and membrane proteins, which are frequently targets of drug development. Once bound to the chip, proteins are highly unstable and their structures often do not reflect the true conformation found under physiological conditions.
Proteins form the important structural and functional machinery of the cell, and are the molecular entities with which nearly all of today's marketed drugs interact. Proteins are thus drug targets. Most pharma companies are investing heavily to extract truly promising drug targets from their sea of unvalidated targets derived from gene-based approaches. Typically the mechanism of action defining how drugs act upon their targets is poorly understood; for some marketed drugs the target is not even known. Furthermore, identifying “non-target” proteins with which the drug interacts to trigger side effects has been especially elusive. It is believed that side effects of many drugs could be diminished with a greater understanding of the mechanism of action involving their target and the non-target proteins.
Drug programs are discontinued for a variety of reasons (e.g., lack of efficacy compared to placebo), but about half of the terminations relate to clinical safety and toxicity. As a result, the developments of many ill-chosen lead drug compounds are halted late in clinical trials after many years and millions of dollars have been spent. Compounding the financial problems caused by toxicity, the long duration of drug development also substantially reduces the length of patent protection.
Adverse side effects from drugs result in more than two million hospitalizations and more than 100,000 deaths each year. Many major drugs have severe toxic side effects.                The widely prescribed psoriasis drugs methoxetrate and cyclosporine can cause severe liver and kidney damage and are thus rarely prescribed for more than one year.        Approximately $13 billion has been spent so far in product injury and class action litigation connected with the withdrawal of the fen-phen weight loss drug combination.        Substantial liabilities were also associated with the hepatotoxicity of the diabetes drug Rezulin (Troglitazone), which was prescribed 2 million times and resulted in 398 deaths before its withdrawal from the market; 8700 law suits are being filed.        Baycol, a cholesterol-lowering statin taken by 700,000 Americans, was removed from the market due to reports of a sometimes fatal muscle-related adverse reaction (rhabdomyolysis) and 31 deaths in the USA. Projected annual Baycol revenues prior to the recall were approximately $1 billion.        Sales growth of Celebrex and Vioxx, blockbusters for the treatment of arthritis, has also been negatively affected by reports of a potential link to heart problems.        
Thus, there is a need to reduce time and costs of drug development by (a) accelerating the hit-to-drug selection process by filtering out those hits likely to trigger side effects and (b) re-engineering drug chemical structure based on the knowledge of drug-target and drug-non-target interactions, reducing or eliminating the undesired interactions.
There is also a need to develop technologies for analysis of the proteome that allow scaling up to industrial levels with the features of an industrial process: high accuracy, reproducibility and flexibility in that the process is high-throughput, automatable and cost-effective. There is a need to develop technologies that permit probing and identification of proteins and other biomolecules in their native conformation using automated protocols and systems therefor. In particular, there is a need to develop strategies and technologies for identification and characterization of hydrophobic proteins under physiological conditions.